Waseda University researchers have demonstrated an optical transmitter capable of sending data at 180 Gbit/s without a conventional optical isolator. The team combined a tunable silicon-photonics laser with a thin-film lithium-niobate modulator and reported a tuning span of approximately 67 nanometers. Waseda publicized the findings in English on November 4, 2026, following an October 7 paper in the Journal of Lightwave Technology. The result is a laboratory demonstration, not a qualified commercial data-center module.
NVIDIA’s official March 18, 2025 overview of co-packaged optical switches. Authentic industry background on photonic networking; NVIDIA hardware is unrelated to Waseda’s experimental transmitter. Credit: NVIDIA / YouTube.
Why a tiny reflection can disrupt a laser
An optical transmitter converts electrical information into modulated light for transmission through fiber. A laser supplies the carrier, while a modulator changes a property of the beam to encode information. Optical isolators normally prevent reflected light from travelling backward into the laser cavity. Without protection, reflections from downstream components can introduce phase noise, unwanted modes or unstable output.
Removing an isolator could simplify assembly and reduce component count, especially as engineers integrate photonic functions onto compact packages. Readers unfamiliar with pluggable data-center optics can consult BitcoinVersus.Tech’s guide to SFP, QSFP and 800G transceivers. Waseda has not reported a packaged transceiver with an Ethernet interface or established fleet reliability.
Two optical platforms, one coordinated response
The design joins a semiconductor gain chip to a silicon-photonics external cavity with two silicon-nitride ring resonators. Together, the rings form a Vernier wavelength filter. The thin-film lithium-niobate section contains an asymmetric Mach–Zehnder interferometer and an electro-optic modulator. Researchers coordinated the spectral response of the filter and interferometer so returning light becomes less disruptive to laser oscillation.
Laboratory measurements showed stable single-mode operation across the wavelength span without an optical isolator or active laser-temperature control. Waseda’s Japanese-language research announcement additionally reports an approximately 0.6 V modulator drive at 180 Gbit/s, with an open eye pattern and without digital signal-processing waveform compensation. A drive-voltage measurement is not the transmitter’s total power consumption, and an open eye pattern does not alone establish production-grade bit-error performance.
NVIDIA’s official August 26, 2025 demonstration of a Quantum-X Photonics co-packaged switch with GB300 racks and OSFP optical modules. Separate production-oriented networking technology, not the Waseda device. Credit: NVIDIA / YouTube.
What 180 Gbit/s means for data centers
At the stated modulation rate, 180 billion bits per second correspond to an ideal 22.5 GB/s before any framing, coding or system overhead. The laboratory result does not establish 180 Gbit/s of application throughput. A tunable laser could eventually support wavelength-division multiplexing, where multiple optical carriers share a fiber, but researchers have not demonstrated an integrated multi-wavelength production link here.
The distinction matters when comparing different achievements. BitcoinVersus.Tech recently covered a 256-wavelength, 81.92 Tb/s transmission experiment; its aggregated multi-channel rate cannot be equated with Waseda’s single-transmitter modulation test. Similarly, ScaleFibre’s 1.6T OSFP announcement describes a different product class and readiness stage.
What must happen before deployment
Packaging, thermal cycling, coupling losses, manufacturability and optical-feedback tolerance under real network conditions still require validation. Operators would also need measured energy per transmitted bit, error-rate performance with practical receivers, compatibility with standardized signaling and service-life data. Published research does not demonstrate all of those requirements or a customer deployment.
The effort nevertheless addresses an important packaging bottleneck: reducing the number of discrete parts while preserving laser stability. Other photonics advances, including lower-loss integrated optical switches, target different limitations of future optical systems. Waseda’s result advances one transmitter architecture; commercial cost, scaling and field reliability remain open questions.
Primary sources and verification
Primary documentation: Waseda University’s November 4, 2026 English release; Waseda’s Japanese technical announcement; Kita and colleagues, Journal of Lightwave Technology, October 7, 2026. Experimental figures originate with the researchers. Independent replication and production qualification were not performed by BitcoinVersus.Tech.
Editor’s Note: The 180 Gbit/s figure is a lab demonstration, not an Ethernet product benchmark. The approximately 0.6 V value describes the modulator drive rather than whole-device power. The original 1200 × 630 photorealistic cover portrays a fictional engineer and laboratory, not Waseda personnel or equipment. Both video recordings are genuine NVIDIA footage illustrating related photonic networking hardware, not Waseda’s experiment. The source image from Waseda has not been republished because its stated restrictions prohibit reuse without permission. Live playback across mobile browsers has not been independently verified.
Support BitcoinVersus.Tech
BitcoinVersus.Tech is volunteer-run. An optional 5,000-satoshi donation supports independent technology research. Scan the original QR from another screen using a compatible Bitcoin wallet. On a phone, save the QR image and import it into a wallet that supports scanning saved pictures. Verify recipient and amount before sending.

Verify recipient 3C9o19EH5HSiwEPyCTmEKzxhNCbo2X6TTb and 5,000 satoshis before approval. Bitcoin payments are generally irreversible. Contributions are entirely voluntary; editorial conclusions are independent. Reporting is educational and informational, not financial advice or certification of engineering performance.

Leave a Reply